Laboratory colonisation and genetic bottlenecks in the tsetse fly Glossina pallidipes

Abstract

Background: The IAEA colony is the only one available for mass rearing of Glossina pallidipes, a vector of human and animal African trypanosomiasis in eastern Africa. This colony is the source for Sterile Insect Technique (SIT) programs in East Africa. The source population of this colony is unclear and its genetic diversity has not previously been evaluated and compared to field populations.

Methodology/principal findings: We examined the genetic variation within and between the IAEA colony and its potential source populations in north Zimbabwe and the Kenya/Uganda border at 9 microsatellites loci to retrace the demographic history of the IAEA colony. We performed classical population genetics analyses and also combined historical and genetic data in a quantitative analysis using Approximate Bayesian Computation (ABC). There is no evidence of introgression from the north Zimbabwean population into the IAEA colony. Moreover, the ABC analyses revealed that the foundation and establishment of the colony was associated with a genetic bottleneck that has resulted in a loss of 35.7% of alleles and 54% of expected heterozygosity compared to its source population. Also, we show that tsetse control carried out in the 1990's is likely reduced the effective population size of the Kenya/Uganda border population.

Conclusions/significance: All the analyses indicate that the area of origin of the IAEA colony is the Kenya/Uganda border and that a genetic bottleneck was associated with the foundation and establishment of the colony. Genetic diversity associated with traits that are important for SIT may potentially have been lost during this genetic bottleneck which could lead to a suboptimal competitiveness of the colony males in the field. The genetic diversity of the colony is lower than that of field populations and so, studies using colony flies should be interpreted with caution when drawing general conclusions about G. pallidipes biology.

Figure 1. Competing scenarios considered in the ABC analysis of IAEA colony past demography (analysis 3).

The demography of the Busia and Rukomeshi populations was determined as described in Supplementary file S2. In all scenarios, the two potential source populations merged T_anc. ago into an unsampled ancestral population and a bottleneck started in Busia in 1991 and lasted BD _cont. generations. Scenario 4, 5 and 6 are respectively variations of scenarios 1, 2 and 3 in which a genetic bottleneck (of duration BD _col.Bus. or BD _col.Ruk.) is associated with the laboratory colonisation. In scenarios 1, 2, 4 and 5 the IAEA colony has a single population of origin. In scenarios 1 and 4, the IAEA colony was founded from the Busia population in 1975 while in scenarios 2 and 5, it was founded from the Rukomeshi population T_col.Ruk. generations ago. In scenario 3 and 6 the IAEA colony originates from an admixture between unsampled colonies of Busia and Rukomeshi origin that were respectively founded from the Busia population in 1975 and from the Rukomeshi population T_col.Ruk. generations ago. When admixture occurs, the admixture rate ar is the proportion from unsampled Busia colony that contributed to the admixed IAEA colony.

Figure 2. Genetic variations within samples.

Error bars indicate the standard deviations across loci. Na: average number of alleles per locus. AR: allelic richness. AR is based on minimum sample size (N = 23 in Friuli for locus DVV-ET1). H: mean expected heterozygosity.

Figure 3. Distribution of the allele frequency classes.

(A) in Busia. (B) in the IAEA colony samples from 2012 and 2013. (C) in Rukomeshi. The result of the bottleneck tests (Wilcoxon' tests on heterozygosity excess, and the mode shift tests) are indicated for each panel which correspond to a field population or to the IAEA colony. Heterozygosity excess and/or a shifted mode in the distribution of allele frequency classes indicate a recent reduction in population size.

Figure 4. Estimated number of clusters and population structure from the Structure analysis.

(A) Mean (±SD) natural logarithm of the likelihood of the data [LnP(X|K)] over 10 Structure replicated runs for each value of the putative number of clusters (K). (B) Estimated population structure from the Structure analysis for K = 2 and 3. Each individual is represented by a vertical line divided into K coloured segments that represent the individual's estimated membership fractions in K clusters. Black lines separate individuals from different samples. Each plot, is based on the highest-probability run (among ten) at K = 2 and 3.

Figure 5. Effective population sizes (Ne) estimated from the ABC analysis with scenario 4 (Figure 1).

The prior distributions are shown as black lines while the posterior distributions are shown as coloured lines using the same colour code as in Figure 1. The medians of a posterior distribution, considered as point estimate of the parameters, are indicated. Each distribution was obtained from 10000 values. (A) N _Bus.: Ne of the Busia population. N _Ruk.: Ne of the Rukomeshi population. N _IAEA.: Ne of the IAEA colony. The prior distributions for N _Bus. and N _Ruk. are shown as a plain line while the prior distribution for N _IAEA. is shown as a dashed line. (B) NB_{col. Bus.}: Ne of the colony of Busia origin (which is at the origin of the current IAEA colony) during the bottleneck associated with its establishment.

Figure 6. Loss of genetic diversity in the IAEA colony (2013 sample) with respect to its simulated source population at the Kenya/Uganda border in 1975.

Na; allelic diversity loss and H; expected heterozygosity loss. Both losses are statistically significant (Wilcoson's signed-rank tests over loci, p≤0.05).